CN110390930A - A kind of method and system of audio text check and correction - Google Patents
A kind of method and system of audio text check and correction Download PDFInfo
- Publication number
- CN110390930A CN110390930A CN201810334974.1A CN201810334974A CN110390930A CN 110390930 A CN110390930 A CN 110390930A CN 201810334974 A CN201810334974 A CN 201810334974A CN 110390930 A CN110390930 A CN 110390930A
- Authority
- CN
- China
- Prior art keywords
- check
- correction
- text
- subfile
- audio
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
Abstract
The embodiment provides a kind of method and system of audio text check and correction, this method includes that audio file is sent to check and correction platform by initiating terminal;Platform is proofreaded by audio file transcription into text file;Platform is proofreaded by audio file and the corresponding audio subfile and text subfile for splitting into no less than 1 of text file;It proofreads platform and audio subfile and corresponding text subfile is sent to check and correction terminal;Terminal display audio subfile and corresponding text subfile are proofreaded, and corresponding text subfile is adjusted according to audio subfile, generates text subfile after check and correction;Check and correction terminal will proofread word subfile hereinafter and be sent to check and correction platform;Check and correction platform merges text subfile after the different check and correction sent of check and correction terminals, the text file after forming check and correction, and is sent to initiating terminal.Using technical solution of the present invention, the working efficiency of audio text check and correction can be effectively improved, and can guarantee the quality of audio text check and correction.
Description
Technical field
The present invention relates to the method and systems that technical field of voice recognition more particularly to a kind of audio text are proofreaded.
Background technique
With the development of speech recognition technology, audio file easily can be converted into text.Relative to audio
File, the memory space that text file does not only take up is smaller, is easy to save, and some special occasions must be by audio file
It is converted into text, forms written document.
However during audio file is converted into text, often it will appear in transcription error, such as audio file
It pronounces indistinctly or the reasons such as homophone word, causes text cannot be corresponding with audio.And in many industries, such as teaching, training
The industries such as instruction, law need text to accomplish with audio complete corresponding, it is therefore desirable to after audio file is converted into text, then root
Text is proofreaded according to audio.
Currently, audio text check and correction be by one side play audio, on one side check text come what is completed, proofread in this way when
Between it is at least equal or even longer with the play time of audio, and if audio transcription text by dialect or the limit of languages
System, it is also necessary to find specific personnel and complete check and correction, the working efficiency that these result in the check and correction of audio text is very low.
Summary of the invention
In view of above-mentioned technical problem, the embodiment of the present invention provides a kind of method and system of audio text check and correction, Neng Gouyou
Effect ground improves the working efficiency of audio text check and correction, and can guarantee the quality of audio text check and correction.
On the one hand the embodiment of the present invention provides a kind of method of audio text check and correction, comprising the following steps:
Audio file is sent to check and correction platform by initiating terminal;
The check and correction platform is by the audio file transcription at text file;
The platform of proofreading is by the audio file and the corresponding audio Ziwen for splitting into no less than 1 of the text file
Part and text subfile;
The audio subfile and corresponding text subfile are sent to check and correction terminal by the check and correction platform;
Audio subfile described in the check and correction terminal display and corresponding text subfile, and according to the audio subfile
Corresponding text subfile is adjusted, text subfile after check and correction is generated;
Text subfile after the check and correction is sent to the check and correction platform by the check and correction terminal;
The check and correction platform merges text subfile after the different check and correction sent of check and correction terminals, after forming check and correction
Text file, and it is sent to the initiating terminal.
Further, further comprising the steps of:
The initiating terminal sends check and correction require information to the check and correction platform, and the check and correction requires to include selection check and correction eventually
End, the languages of transcription text file, text file exhibition method, feedback time and/or secrecy.
Further, further comprising the steps of:
The initiating terminal consults check and correction progress by the check and correction platform;
The initiating terminal sends the instruction for abandoning proofreading or reselecting check and correction terminal to the check and correction platform, described
Check and correction platform is abandoned proofreading or reselecting check and correction terminal according to instruction;
Or the initiating terminal evaluates different check and correction terminals.
Further, further comprising the steps of:
The audio file that the check and correction platform sends the initiating terminal carries out quality testing, judgement can transcription or
Check and correction.
Further, the check and correction platform splits into no less than 1 to the audio file and text file correspondence
Audio subfile and text subfile, further include steps of
The audio file is split into no less than 1 audio subfile according to fixed duration by the check and correction platform, and
The text file is accordingly split into no less than 1 text subfile.
Further, the check and correction platform splits into no less than 1 to the audio file and text file correspondence
Audio subfile and text subfile, further include steps of
The text file is split into no less than 1 text subfile according to whole sentence by the check and correction platform, and corresponding
The audio file is split into no less than 1 audio subfile by ground.
Further, further comprising the steps of:
The check and correction platform evaluates different initiating terminals and/or different check and correction terminals.
Further, further comprising the steps of:
The check and correction terminal evaluates the check and correction platform and/or different initiating terminals, and obtains check and correction expense.
On the other hand the embodiment of the present invention additionally provides a kind of system of audio text check and correction, including initiating terminal, school
To platform and check and correction terminal, the initiating terminal, the check and correction platform and the check and correction terminal pass through network connection, wherein
The initiating terminal is used to for audio file being sent to the check and correction platform, and obtains and proofread from the check and correction platform
Text file afterwards;
The check and correction platform is used to obtain the audio file from the initiating terminal, and the audio file transcription is written
Word file, by the audio file and corresponding audio subfile and the text Ziwen for splitting into no less than 1 of the text file
The audio subfile and corresponding text subfile are sent to check and correction terminal, and the school that different check and correction terminals are sent by part
Rear text subfile is merged, the text file after forming check and correction is sent to the initiating terminal;
The check and correction terminal is used to show the audio subfile obtained from the check and correction platform and corresponding text
File adjusts corresponding text subfile according to the audio subfile, generates text subfile after check and correction, and by the check and correction
Text subfile is sent to the check and correction platform afterwards.
Further, the check and correction platform further comprises split cells, and the split cells is used for according to fixed duration
The audio file is split into no less than 1 audio subfile, and the text file is accordingly split into no less than 1
A text subfile, or the text file is split into according to whole sentence no less than 1 text subfile, and accordingly
The audio file is split into no less than 1 audio subfile.
Above-mentioned technical proposal have the following advantages that or the utility model has the advantages that
1, by splitting audio file and corresponding text file, allow audio text proof-reading parallel
Carry out, to reduce operating time, improves work efficiency.
2, due to can mutually be evaluated between the promoter of audio text check and correction, check and correction platform and proofreader, so as to
Accomplish to appreciate it is excellent penalize bad, mutually promote, improve the quality of proofreading of audio text.
3, due to the addition of the proofreader largely with different abilities, the special money of audio text check and correction can be obtained quickly
Source expands the coverage area of audio text check and correction.
4, since the promoter of audio text check and correction can independently select proofreader to provide check and correction task by proofreading platform,
Therefore privacy requirement and the time requirement of promoter be can satisfy.
Detailed description of the invention
Fig. 1 is the flow chart of one sound intermediate frequency text of embodiment of the present invention check and correction.
Fig. 2 is the structural schematic diagram of two sound intermediate frequency text critique system of the embodiment of the present invention.
Specific embodiment
The exemplary embodiment that the present invention will be described in more detail below with reference to accompanying drawings.Although showing the present invention in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the present invention without should be by embodiments set forth here
It is limited.It is to be able to thoroughly understand the present invention on the contrary, providing these embodiments, and can be by the scope of the present invention
It is fully disclosed to those skilled in the art.
Fig. 1 is the flow chart of one sound intermediate frequency text of embodiment of the present invention check and correction.As shown in Figure 1, audio text check and correction
Process the following steps are included:
The promoter that step 101, audio text proofread task obtains audio file on initiating terminal, which can
To be computer, it is also possible to mobile phone, audio file can be obtained from other equipment by network, can also locally passing through record
Sound obtains audio file, and promoter sends audio text to check and correction platform by initiating terminal and proofreads mission bit stream.
The mission bit stream includes audio file itself, further includes the require information for audio file check and correction task, check and correction
Require information can be selection check and correction terminal, namely select which personnel to proofread to the task of oneself, such as internal populations,
Or ordinary person on platform;It can be the languages of transcription text file, and whether be related to specialized vocabulary, need specific
Personnel could complete;It can also be the last exhibition method of text file, such as subtitle or document;It can also be and require task
Feedback time, if need urgent;It is also possible to security requirements, etc..
Step 102, check and correction platform receive the audio text check and correction mission bit stream that initiating terminal is sent, and first check for therein
Audio file quality, judgement can transcription or check and correction, then by the audio file transcription at text file.
During transcription, platform is proofreaded according to the requirement of promoter, audio file can be submitted to different engines
Transcription, such as domestic urgent task, can submit to the speech transcription engine supplier of check and correction platform cooperation, domestic free
Task, other speech transcription engines supplier that can submit to check and correction platform cooperation can submit to overseas task
The speech transcriptions engine such as IBM or GOOGLE.
In this step, check and correction platform can also estimate the cost time of the task.And by the task order shape of promoter
State feeds back to the initiating terminal of promoter.
Step 103, check and correction platform by audio file and text file it is corresponding split into 1 or multiple audio subfiles and
Text subfile.
There are many modes of fractionation, first way can be according to fixed duration by audio file split into 1 or
Multiple audio subfiles, such as 1 minute, 2 minutes etc., then text file is accordingly split into the text Ziwen of identical quantity
Part.
The second way is text file to be split into 1 or multiple text subfiles according to whole sentence, and accordingly will
Audio file splits into the audio subfile of identical quantity, and this fractionation mode can guarantee that each subfile has been semantically
Whole.
After splitting, each audio subfile text subfile corresponding with its should be consistent in terms of content.
Having split for task is shown, and notifies the check and correction terminal of proofreader by step 104, check and correction platform, there is new post
Business order.Audio subfile and corresponding text subfile are sent to check and correction terminal or the proofreader of proofreader by check and correction platform
It is obtained from check and correction platform by check and correction terminal.
It is whole in check and correction after step 105, proofreader obtain audio subfile and corresponding text subfile by check and correction terminal
End plays audio subfile, opens simultaneously corresponding text subfile, and adjust corresponding text according to the content of audio subfile
Word subfile, to generate text subfile after check and correction.
Step 106, proofreader are sent to check and correction platform by text subfile after proofreading the check and correction that terminal completes oneself.
Step 107, check and correction platform merge text subfile after the different check and correction sent of check and correction terminals, form check and correction
Text file afterwards, and it is sent to initiating terminal, promoter obtains the text file after check and correction.
In above process, promoter can consult check and correction progress by check and correction platform by initiating terminal, can also be with
The instruction for abandoning proofreading or reselecting proofreader's (check and correction terminal) is sent to check and correction platform, check and correction platform is abandoned according to instruction
Proofread or reselect proofreader (check and correction terminal), can also by initiating terminal to different proofreader'ss (check and correction terminal) into
Row evaluation marking, high for evaluating, promoter can carry out beating reward to proofreader (check and correction terminal).
Different promoter'ss (initiating terminal) and/or different proofreader (check and correction terminal) can also be carried out by proofreading platform
Evaluation.For promoter (initiating terminal), processing, the example such as alert or degrade primarily with respect to the malicious act of promoter
It is such as poor to the malice of proofreader comment, task order malice it is poor comment, malice extension task dispatching.It is main for proofreader (check and correction terminal)
If carrying out bonus point or deduction strategy to proofreader according to the feedback of promoter, for having passed through the task of audit, can incite somebody to action
Task distribution of bonus is to proofreader.
Check and correction terminal can also evaluate check and correction platform and/or different promoter (initiating terminal), and obtain school
To expense.
In order to realize above-mentioned technical proposal, on the other hand the embodiment of the present invention additionally provides a kind of audio text check and correction
System, Fig. 2 are the structural schematic diagram of two sound intermediate frequency text critique system of the embodiment of the present invention.
As shown in Fig. 2, the system of audio text check and correction includes initiating terminal 201, check and correction platform 202 and check and correction terminal
203, check and correction platform further comprises split cells 204, and initiating terminal, check and correction platform and check and correction terminal pass through network connection.
Initiating terminal is audio file to be sent to check and correction platform, and the text file after check and correction is obtained from check and correction platform.
It proofreads platform and obtains audio file from initiating terminal, by audio file transcription at text file.
Split cells in check and correction platform, which corresponds to audio file and text file, splits into 1 or multiple audios
Audio file can be split into 1 or multiple sounds according to fixed duration in split cells by file and text subfile
Frequency subfile, and accordingly by text file split into 1 perhaps multiple text subfile or according to whole sentence by text text
Part splits into 1 and perhaps multiple text subfiles and audio file is accordingly split into 1 or multiple audio Ziwens
Part.
Audio subfile and corresponding text subfile are also sent to check and correction terminal by check and correction platform, and eventually by different check and correction
Text subfile after the check and correction sent is held to merge, the text file after forming check and correction is sent to initiating terminal.
Audio subfile and corresponding text subfile that terminal display is obtained from check and correction platform are proofreaded, according to audio Ziwen
Part adjusts corresponding text subfile, generates text subfile after check and correction, and will check and correction to be sent to check and correction flat for word subfile hereinafter
Platform.
It can mutually be evaluated between initiating terminal, check and correction platform and check and correction terminal.Promoter can be logical by initiating terminal
Check and correction platform is crossed to consult check and correction progress, can also abandon proofreading or reselecting proofreader's (check and correction to check and correction platform transmission
Terminal) instruction, check and correction platform abandons proofreading or reselect proofreader (check and correction terminal) according to instruction, can also pass through hair
It plays terminal and evaluation marking is carried out to different proofreader'ss (check and correction terminal), high for evaluating, promoter can be to proofreader (school
To terminal) it carries out beating reward.
Different promoter'ss (initiating terminal) and/or different proofreader (check and correction terminal) can also be carried out by proofreading platform
Evaluation.For promoter (initiating terminal), processing, the example such as alert or degrade primarily with respect to the malicious act of promoter
It is such as poor to the malice of proofreader comment, task order malice it is poor comment, malice extension task dispatching.It is main for proofreader (check and correction terminal)
If carrying out bonus point or deduction strategy to proofreader according to the feedback of promoter, for having passed through the task of audit, can incite somebody to action
Task distribution of bonus is to proofreader.
Check and correction terminal can also evaluate check and correction platform and/or different promoter (initiating terminal), and obtain school
To expense.
Using above-mentioned technical proposal, by splitting audio file and corresponding text file, so that audio is literary
Word proof-reading can be carried out parallel, to reduce operating time, improve work efficiency;Due to the hair of audio text check and correction
Rising can mutually evaluate between people, check and correction platform and proofreader, so as to accomplish to appreciate it is excellent penalize bad, mutually promote, improve audio
The quality of proofreading of text;Due to the addition of the proofreader largely with different abilities, the check and correction of audio text can be obtained quickly
Special resource expands the coverage area of audio text check and correction;Since the promoter of audio text check and correction can be flat by proofreading
Platform independently selects proofreader to provide check and correction task, therefore can satisfy privacy requirement and the time requirement of promoter.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that the independent physics of each unit includes, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of hardware adds SFU software functional unit.
The above-mentioned integrated unit being realized in the form of SFU software functional unit can store and computer-readable deposit at one
In storage media.Above-mentioned SFU software functional unit is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) executes receiving/transmission method described in each embodiment of the present invention
Part steps.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (Read-Only Memory, abbreviation
ROM), random access memory (Random Access Memory, abbreviation RAM), magnetic or disk etc. are various can store
The medium of program code.
Above-described is the preferred embodiment of the present invention, it should be pointed out that the ordinary person of the art is come
It says, several improvements and modifications can also be made under the premise of not departing from principle of the present invention, these improvements and modifications also exist
In protection scope of the present invention.
Claims (10)
1. a kind of method of audio text check and correction, which comprises the following steps:
Audio file is sent to check and correction platform by initiating terminal;
The check and correction platform is by the audio file transcription at text file;
The check and correction platform by the corresponding audio subfile for splitting into no less than 1 of the audio file and the text file and
Text subfile;
The audio subfile and corresponding text subfile are sent to check and correction terminal by the check and correction platform;
Audio subfile described in the check and correction terminal display and corresponding text subfile, and adjusted according to the audio subfile
Corresponding text subfile generates text subfile after check and correction;
Text subfile after the check and correction is sent to the check and correction platform by the check and correction terminal;
The check and correction platform merges text subfile after the different check and correction sent of check and correction terminals, the text after forming check and correction
File, and it is sent to the initiating terminal.
2. a kind of method of audio text check and correction according to claim 1, which is characterized in that further comprising the steps of:
The initiating terminal to the check and correction platform send check and correction require information, it is described check and correction require to include selection check and correction terminal,
Languages, text file exhibition method, feedback time and/or the secrecy of transcription text file.
3. a kind of method of audio text check and correction according to claim 1, which is characterized in that further comprising the steps of:
The initiating terminal consults check and correction progress by the check and correction platform;
The initiating terminal sends the instruction for abandoning proofreading or reselecting check and correction terminal, the check and correction to the check and correction platform
Platform is abandoned proofreading or reselecting check and correction terminal according to instruction;
Or the initiating terminal evaluates different check and correction terminals.
4. a kind of method of audio text check and correction according to claim 1, which is characterized in that further comprising the steps of:
The check and correction platform carries out quality testing to the audio file that the initiating terminal is sent, and can judgement transcription or school
It is right.
5. a kind of method of audio text check and correction according to claim 1, which is characterized in that the check and correction platform is to described
Audio file and the corresponding audio subfile and text subfile for splitting into no less than 1 of the text file further comprise
Following steps:
The audio file is split into no less than 1 audio subfile according to fixed duration by the check and correction platform, and corresponding
The text file is split into no less than 1 text subfile by ground.
6. a kind of method of audio text check and correction according to claim 1, which is characterized in that the check and correction platform is to described
Audio file and the corresponding audio subfile and text subfile for splitting into no less than 1 of the text file further comprise
Following steps:
The text file is split into no less than 1 text subfile according to whole sentence by the check and correction platform, and accordingly will
The audio file splits into no less than 1 audio subfile.
7. a kind of method of audio text check and correction according to claim 1, which is characterized in that further comprising the steps of:
The check and correction platform evaluates different initiating terminals and/or different check and correction terminals.
8. a kind of method of audio text check and correction according to claim 1, which is characterized in that further comprising the steps of:
The check and correction terminal evaluates the check and correction platform and/or different initiating terminals, and obtains check and correction expense.
9. a kind of system of audio text check and correction, which is characterized in that described including initiating terminal, check and correction platform and check and correction terminal
Initiating terminal, the check and correction platform and the check and correction terminal pass through network connection, wherein
The initiating terminal is used to for audio file being sent to the check and correction platform, and after check and correction platform acquisition check and correction
Text file;
The check and correction platform is used to obtain the audio file from the initiating terminal, by the audio file transcription at text text
Part will by the audio file and the corresponding audio subfile and text subfile for splitting into no less than 1 of the text file
The audio subfile and corresponding text subfile are sent to check and correction terminal, and the check and correction that different check and correction terminals are sent is hereinafter
Word subfile merges, and the text file after forming check and correction is sent to the initiating terminal;
The check and correction terminal is used to show the audio subfile and corresponding text subfile obtained from the check and correction platform,
Corresponding text subfile is adjusted according to the audio subfile, generates text subfile after check and correction, and hereinafter by the check and correction
Word subfile is sent to the check and correction platform.
10. a kind of system of audio text check and correction according to claim 9, which is characterized in that the check and correction platform is into one
Step includes split cells, and the split cells is used to split into the audio file according to fixed duration no less than 1 sound
Frequency subfile, and the text file is accordingly split into no less than 1 text subfile, or will be described according to whole sentence
Text file splits into no less than 1 text subfile, and the audio file is accordingly split into no less than 1 sound
Frequency subfile.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810334974.1A CN110390930A (en) | 2018-04-15 | 2018-04-15 | A kind of method and system of audio text check and correction |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810334974.1A CN110390930A (en) | 2018-04-15 | 2018-04-15 | A kind of method and system of audio text check and correction |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110390930A true CN110390930A (en) | 2019-10-29 |
Family
ID=68282577
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810334974.1A Pending CN110390930A (en) | 2018-04-15 | 2018-04-15 | A kind of method and system of audio text check and correction |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110390930A (en) |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101539929A (en) * | 2009-04-17 | 2009-09-23 | 无锡天脉聚源传媒科技有限公司 | Method for indexing TV news by utilizing computer system |
CN103345922A (en) * | 2013-07-05 | 2013-10-09 | 张巍 | Large-length voice full-automatic segmentation method |
CN105161094A (en) * | 2015-06-26 | 2015-12-16 | 徐信 | System and method for manually adjusting cutting point in audio cutting of voice |
CN105245917A (en) * | 2015-09-28 | 2016-01-13 | 徐信 | System and method for generating multimedia voice caption |
CN105302795A (en) * | 2015-11-11 | 2016-02-03 | 河海大学 | Chinese text verification system and method based on Chinese vague pronunciation and voice recognition |
CN106744111A (en) * | 2016-12-05 | 2017-05-31 | 深圳市图焌科技有限公司 | Elevator repair and maintenance management system and method based on bluetooth and Mobile Data Communication Technology |
CN107147564A (en) * | 2017-05-09 | 2017-09-08 | 胡巨鹏 | Real-time speech recognition error correction system and identification error correction method based on cloud server |
CN107220228A (en) * | 2017-06-13 | 2017-09-29 | 深圳市鹰硕技术有限公司 | One kind teaching recorded broadcast data correction device |
CN107820112A (en) * | 2017-11-15 | 2018-03-20 | 安徽声讯信息技术有限公司 | A kind of audio written broadcasting live system |
-
2018
- 2018-04-15 CN CN201810334974.1A patent/CN110390930A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101539929A (en) * | 2009-04-17 | 2009-09-23 | 无锡天脉聚源传媒科技有限公司 | Method for indexing TV news by utilizing computer system |
CN103345922A (en) * | 2013-07-05 | 2013-10-09 | 张巍 | Large-length voice full-automatic segmentation method |
CN105161094A (en) * | 2015-06-26 | 2015-12-16 | 徐信 | System and method for manually adjusting cutting point in audio cutting of voice |
CN105245917A (en) * | 2015-09-28 | 2016-01-13 | 徐信 | System and method for generating multimedia voice caption |
CN105302795A (en) * | 2015-11-11 | 2016-02-03 | 河海大学 | Chinese text verification system and method based on Chinese vague pronunciation and voice recognition |
CN106744111A (en) * | 2016-12-05 | 2017-05-31 | 深圳市图焌科技有限公司 | Elevator repair and maintenance management system and method based on bluetooth and Mobile Data Communication Technology |
CN107147564A (en) * | 2017-05-09 | 2017-09-08 | 胡巨鹏 | Real-time speech recognition error correction system and identification error correction method based on cloud server |
CN107220228A (en) * | 2017-06-13 | 2017-09-29 | 深圳市鹰硕技术有限公司 | One kind teaching recorded broadcast data correction device |
CN107820112A (en) * | 2017-11-15 | 2018-03-20 | 安徽声讯信息技术有限公司 | A kind of audio written broadcasting live system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10726833B2 (en) | System and method for rapid customization of speech recognition models | |
CN104239459B (en) | voice search method, device and system | |
EP2317508B1 (en) | Grammar rule generation for speech recognition | |
US20060004570A1 (en) | Transcribing speech data with dialog context and/or recognition alternative information | |
Pobar et al. | Online speaker de-identification using voice transformation | |
CN109325091B (en) | Method, device, equipment and medium for updating attribute information of interest points | |
EP1901283A2 (en) | Automatic generation of statistical laguage models for interactive voice response applacation | |
CN105185379A (en) | Voiceprint authentication method and voiceprint authentication device | |
CA2653973A1 (en) | Replacing text representing a concept with an alternate written form of the concept | |
US9412359B2 (en) | System and method for cloud-based text-to-speech web services | |
US20130253932A1 (en) | Conversation supporting device, conversation supporting method and conversation supporting program | |
CN103605492B (en) | A kind of self adaptation speech training method and platform | |
US20110144990A1 (en) | Rating speech naturalness of speech utterances based on a plurality of human testers | |
CN109559749A (en) | Combined decoding method and system for speech recognition system | |
CA2590739A1 (en) | Method and apparatus for voice message editing | |
CN109545203A (en) | Audio recognition method, device, equipment and storage medium | |
US7428491B2 (en) | Method and system for obtaining personal aliases through voice recognition | |
CN106297766A (en) | Phoneme synthesizing method and system | |
US7853451B1 (en) | System and method of exploiting human-human data for spoken language understanding systems | |
CN110390930A (en) | A kind of method and system of audio text check and correction | |
CN110473524A (en) | The construction method and device of speech recognition system | |
KR102474690B1 (en) | Apparatus for taking minutes and method thereof | |
Richmond et al. | Festival multisyn voices for the 2007 blizzard challenge. | |
KR20190070682A (en) | System and method for constructing and providing lecture contents | |
CN105890612A (en) | Voice prompt method and device in navigation process |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20191029 |
|
WD01 | Invention patent application deemed withdrawn after publication |